TERTIARY PROTEACEAE IN AUSTRALIA: A RE-INVESTIGATION OF 
BANKSIA ADUNCA AND DRYANDRA URNIFORMIS 
Robert S. Hill 
Department of Plant Science, University of Tasmania, GPO Box 252C, Hobart, Tasmania 7001 
Hill, Robert S., 1990:05:31. Tertiary Proteaceae in Australia: a re- investigation of 
Banksia adunca and Dryandra urniformis. Proceedings of the Royal Society of Victoria 
102 (1): 23-28. ISSN 0035-921 1. 
The specimens described by Deane (1925) as Banksia adunca and Dryandra urniformis 
from Upper Oligocene-Lower Miocene coals at Morwell, Victoria have been re-examined 
and thcircuticular morphology examined for the first time. Banksia adunca is considered to 
be conspecific with Banksieaephyllum fastigatum (Dean) Cookson & Duigan. D. urniformis 
is also transferred to Banksieaephyllum and is considered to be closely related to B. elon¬ 
gation Hill & Christophel, 1988 from the Upper Oligoccnc-Lowcr Miocene at Loy Yang. 
These Banksieaephyllum species are particularly important in tracing the history of the 
development of sclcrophylly and xcromorphy in Australian plants. 
MACROFOSSILS of the Proteaceae are well 
represented in Australian Tertiary sediments, 
leaves and reproductive structures of a range of 
genera having been described (Cookson & 
Duigan 1950, Pike 1953, Lange 1978, Blackburn 
1981, McNamara & Scott 1983, Christophel 
1984, Carpenter & Hill 1988, Hill & Christophel 
1988). Amongst the Proteaceae the most com¬ 
mon and diverse macrofossils belong to the tribe 
Banksieae. 
The coals of the Latrobe Valley in Gippsland, 
Victoria contain a diversity of macrofossils in 
the Banksieae, nine species based on vegetative 
material having been described by Deane 
(1925), Cookson & Duigan (1950) and Hill & 
Christophel (1988). Following Ettingshausen 
(1888), Deane (1925) adopted the convention of 
placing serrate leaves in Banksia and pinnate 
leaves in Dryandra but he was aware that this 
distinction is invalid amongst living species. 
Cookson & Duigan (1950) concluded that extant 
species of these two genera cannot be separated 
on leaf form alone and thus erected the form 
genus Banksieaephyllum for fossil leaves with 
the characteristics of Banksia and/or Dryandra. 
They transferred Banksia fastigata Deane to 
Banksieaephyllum on the basis of the original 
specimen and new collections. The other two 
species described by Deane, Banksia adunca and 
Dryandra urniformis , were listed by Cookson & 
Duigan amongst species requiring re-investi¬ 
gation to establish their affinities. Hill & 
Christophel (1988) noted that leaves should not 
be placed in Banksieaephyllum unless the 
cuticular morphology was sufficiently well 
preserved to support the assignment. 
The type specimens of Banksia adunca and 
Dryandra urniformis were examined recently in 
the Museum of Victoria (specimen numbers 
with prefix NMV), and it was determined that in 
both specimens the cuticle was preserved. The 
specimens have been re-investigated and their 
cuticular morphology studied to determine 
whether their placement in the tribe Banksieae is 
valid, and if so whether they should be trans¬ 
ferred to the form genus Banksieaephyllum. 
Both fossil species were collected from the 
Morwell coal (Deane 1925) which spans the 
Upper Oligocene to Lower Miocene (Stover & 
Partridge 1973), but the precise horizon from 
which the specimens came is uncertain. 
MATERIAL AND METHODS 
The material of each species consists of only part 
of the leaf. Extreme care was required in prepar¬ 
ing the cuticle since only a small amount of leaf 
tissue is available. Before any attempt was made 
to remove leaf fragments, the specimens were 
photographed using an Olympus OM-4 camera 
with bellows and reflected light. Both leaves 
appear to have been coated with some type of 
varnish for protection, and this was successfully 
removed by soaking small fragments of the 
leaves in acetone for several hours. Following 
this treatment the leaf fragments were placed in 
10% chromic acid until the internal leaf tissue 
had dissolved. The cuticular fragments were 
then washed, neutralised in 5% ammonia, 
stained in 1% aqueous safranin O, and mounted 
on microscope slides in phenol glycerine jelly. 
Cuticles were examined using a Zeiss Axioskop 
compound microscope. 
23 
